Computer-implemented method and device for determining one or more operating loads of a vehicle, in particular a rail vehicle

Abstract

The present invention relates to a computer-implemented method for determining one or more operating loads of a vehicle, in particular a rail vehicle, the method comprising the steps: - generating and / or providing (100) a model of the vehicle; - providing (101) at least one piece of vehicle information of the vehicle by an information providing unit (204); - Determining (110) the one or more operating loads of the vehicle by an operating load determination unit (210) based on the at least one provided piece of vehicle information, wherein artificial intelligence is used to determine the one or more operating loads. The invention further relates to a device (200) for carrying out the method.

Description

The present invention relates to a computer-implemented method and to a device for determining one or more operating loads of a vehicle, in particular of a rail vehicle.The service life of vehicles and in particular of their components usually depends on a number of factors. Such factors may include, but are not limited to, the material used, the design, and / or the number and / or strength of (mechanical) influences.Especially in the construction of rail vehicles, attention is increasingly paid to lightweight construction. New materials are used to meet the foregoing challenges. However, this presupposes a design of the respective components.At present, the structural components, in particular in the case of rail vehicles, are usually designed on the basis of standards and / or standard values. However, this has the disadvantage that the components designed therewith are generally oversized.Furthermore, rail vehicles are exposed during operation to loads which are not covered or taken into account by the known standards. For example, rail vehicles are exposed to vibrations and / or loads on account of the rail geometry and unevenness in the track, which vibrations and / or loads are not taken into account in the existing standards.Known design models, such as multi-body models, are also time- and cost-intensive. In addition, such models are very complicated to create. A central role for the determination of the aforementioned loads and for the design of components is the determination of so-called operating loads, i.e. (mechanical) loads which occur during operation of the vehicle.The object of the invention is therefore to specify a method and a device with the aid of which one or more operating loads of a vehicle can be determined easily and accurately.With regard to the method, the object is achieved according to the invention by a computer-implemented method according to the independent method claim. With regard to the device, the object is achieved according to the invention by a device according to the independent device claim.Advantageous embodiments, developments and variants are the subject matter of the dependent claims.The advantages and preferred embodiments listed with regard to the method are to be transferred analogously to the device and vice versa.Specifically, the object directed to the method is achieved by a computer-implemented method for determining one or more operating loads of a vehicle.The vehicle is in particular-but not restrictive-a rail vehicle. The rail vehicle can be, for example, a passenger train or goods train.The described method and the described device can, however, generally also be transferred to other types of vehicles, such as a truck or a watercraft.Preferably, the method comprises generating and / or providing a model of the vehicle.Preferably, a model already created during the construction of the vehicle can be used.Alternatively or additionally, a new model of the vehicle is preferably generated. Geometric details of the vehicle can be used here, for example.Furthermore, at least one piece of vehicle information of the vehicle is preferably subsequently provided.The at least one piece of vehicle information is preferably provided by an information provision unit.It may be advantageous if a plurality of items of vehicle information are provided. This preferably increases the accuracy of the method.The at least one piece of vehicle information can be, for example, but not by way of limitation, technical information and / or operating information about the vehicle and / or about a component of the vehicle.As a further step of the method, the one or more operating loads of the vehicle are preferably ascertained by an operating load ascertainment unit. If an operating load is referred to in a simplified manner below, a plurality of operating loads are also always meant here, unless the opposite is explicitly mentioned.It can be advantageous if the one or more operating loads are determined on the basis of the at least one provided vehicle information item.According to the invention, an artificial intelligence is advantageously used to determine the one or more operating loads.The advantage here is that by means of the artificial intelligence (hereinafter also referred to simply as AI) it is also possible to determine operating loads of and / or in and / or on the vehicle and / or components of the vehicle, via which no vehicle information is present.The AI is preferably programmed and / or trained to determine an overall determination of one or more operating loads. The term "overall determination" can preferably be understood in the sense of this application to mean that the one or more determined operating loads provide sufficient information about the loads occurring on the entire vehicle.Preferably, the input data of the AI is the at least one piece of vehicle information. Further preferably, the output data of the AI is the operation load.In one specific embodiment, a numerical method is used for generating and / or providing the model of the vehicle.It can be advantageous if the model of the vehicle is a finite element model. Preferably, the finite element model is based on a calculation of a nonlinear equation system, which is created on the basis of the input data. Output data may be, for example, but not limited to, deformation, strain, and / or (mechanical) tension of the vehicle.Such models have preferably proven to be expedient for ascertainments of the above type.The at least one piece of vehicle information can preferably be taken into account when creating and / or providing the model.However, the model of the vehicle is not limited to a numerical method, and particularly, to a finite element model. Rather, in an alternative embodiment, other model types can also be used.In one embodiment, the information providing unit comprises a database. In an alternative embodiment, the information provision unit is designed as a database.The at least one piece of vehicle information is preferably provided by the database.According to an alternative embodiment, the database can also be an external database. The term "external database" can be understood in the sense of this application to mean that the database is preferably not part of the information provision unit.In this case, the information provision unit can preferably access data of the database by means of a data connection. The data connection is preferably configured to be wired or wireless.It can be advantageous if the database is created when the model of the vehicle is generated and / or provided.Particularly preferably, the database is created automatically. It can be advantageous if the automatic creation of the database takes place by means of a specially coded program.For example, the specially coded program can be a python code.During the automatic creation of the database, a finite element model is preferably first created by a user. The program can then, preferably based on parameters defined by the user, preferably generate random inputs and start one or more simulations. Advantageously, the program can then extract results useful for the database and store them in the database.A general and non-limiting example will now be explained to illustrate the automatic creation of the database:A force is applied to an object. A finite element model is created based on the object and the applied force. The program, e.g., the python code, creates various models with a random force that are within a range defined by the user. The program then starts the simulation and retrieves the simulation results, which are subsequently stored in the database.In one embodiment, the database contains vehicle information which has been recorded and / or created, for example, during tests and / or in all-day operation of the vehicle.In one embodiment, the generation and / or provision of the model of the vehicle takes place by one or more of the following steps:inputting a number of desired simulations; and / orinput of an effect to be simulated on the vehicle; and / orinputting a number of outputs of the model to be called.The input is preferably performed by a user. The user inputs, for example, how many models (=Simulationen) are desired, which parameters vary (e.g., an applied force) and which and / or how many results (=Ausgaben) are to be obtained.The number of simulations desired may preferably depend on the complexity of the model and / or of the vehicle. For example, it is expedient to carry out a higher number of simulations in the case of a complex vehicle or a complex component than in the case of a simple vehicle or a simple component.A higher number of simulations preferably enables a high resolution of the model. Thus, the model advantageously enables an accurate determination of the operating loads.The term "effect on the vehicle" can be understood in the sense of this application preferably to mean a mechanical effect on the vehicle. For example, the effect is a force amplitude of a force acting on the vehicle or the component.However, the application is not limited to the above force amplitude. Instead, other, in particular mechanical, influences on the vehicle can also be input.It can be advantageous if several influences to be simulated are input.Preferably, an input of a region of an action to be simulated takes place. For example, and with respect to the aforementioned force amplitude, the region of the action to be simulated can be a range of values of acting forces that are to be simulated.In one embodiment, the information provision unit has at least one sensor. Preferably, the information providing unit comprises a plurality of sensors.The at least one or more sensors are preferably arranged on and / or in the vehicle.The at least one piece of vehicle information is preferably provided by the at least one sensor. That is to say that the detected sensor signals are preferably transmitted to the information provision unit.In order to provide the at least one piece of vehicle information, an evaluation of the detected sensor signals is particularly preferably carried out by the information provision unit.The evaluation of the detected sensor signals preferably enables an increased level of detail of the at least one piece of vehicle information.Alternatively, the provision of the at least one piece of vehicle information is preferably effected without an, in particular preceding, evaluation of the detected sensor signals.In a preferred embodiment, a plurality of vehicle information is provided by the at least one or more sensors. One and the same detected sensor signal can be used to provide two or more items of vehicle information, which are preferably based on the detected sensor signal.For example, a running performance of the vehicle detected by at least one sensor can be used on the one hand to indicate an (imminent) maintenance deadline and on the other hand to indicate wear.According to an alternative or supplementary embodiment, the information provision unit uses one or more sensors of the vehicle to provide the at least one piece of vehicle information. This means that sensors already present in and / or on the vehicle are preferably used to provide the at least one piece of vehicle information.The advantage of this alternative or supplementary embodiment is that fewer additional sensors are required, which reduces costs and outlay.It can be particularly advantageous if the at least one sensor is one of the following sensors:force sensor; and / oracceleration sensor; and / orshock absorbers; and / orstrain sensor; and / orpressure sensor.The force sensor can preferably be a sensor which detects a force acting on the vehicle or on a component of the vehicle.The acceleration sensor is preferably a sensor which detects negative and / or positive accelerations of the vehicle.The shock absorber is preferably a shock indicator which detects, in particular, shocks and / or impacts of the vehicle. The shock absorber is particularly preferably a multi-axis shock indicator which can detect shocks and impacts in a plurality of axes, that is to say directions. Alternatively, the shock absorber may also be a single-axis shock indicator.Preferably, but not by way of limitation, the strain sensor is a sensor which detects a deformation, in particular a strain, of the vehicle or of a component of the vehicle. For example, but not by way of limitation, the strain sensor is a strain gauge.The pressure sensor can preferably be a sensor which detects a pressure acting on the vehicle or on a component of the vehicle.Information about the above-described mechanical effects on the vehicle or on a component of the vehicle is advantageous since this preferably allows information about the use of a specific material and / or a material thickness at the location at which the at least one sensor is placed to be derived.According to one embodiment, the model of the vehicle is validated. The validation is preferably carried out by means of the at least one piece of vehicle information.The at least one piece of vehicle information is preferably-but not restrictive-the piece of vehicle information detected by means of the at least one sensor.Particularly preferably, during the validation of the model, a comparison of the sensor data with simulation results takes place.In a particularly preferred embodiment, the AI is trained with the at least one provided vehicle information.Preferably, the AI is trained by data pairs. In this case, the data pairs can contain, for example and with reference to the sensor types mentioned above by way of example, a force value and an associated expansion of a component of the vehicle. However, the data pairs are not limited to the above example.In one embodiment, the AI is trained by supervised learning (supervised learning).Preferably, the AI receives a known data set from inputs and outputs. The AI then preferably adjusts its algorithm to patterns in the data. An operator can advantageously make corrections to optimize the AI.According to another embodiment, the AI is trained by unsupervised learning (Unuputrified Learning).In this case, the AI preferably independently investigates the data and attempts to identify trends and patterns. As the term already indicates, preferably no correction is carried out by an operator.In an alternative embodiment, the AI is trained by semi-supervised learning (semi-supervised learning).Preferably, the AI receives marked and unlabeled data from which it recognizes patterns. The partially supervised learning preferably represents a mixed form of the two above-mentioned training methods.In a particularly preferred embodiment, the AI is designed as an artificial neural network. The advantage of an artificial neural network is that such Kl types can preferably recognize complex patterns in large amounts of data.For this reason, the artificial neural network is preferably designed and configured in such a way that it can be trained by means of the at least one piece of vehicle information and / or to recognize corresponding patterns from the at least one piece of vehicle information, so that more precise operating loads can be derived and / or predicted therefrom.In a simplest configuration, artificial neural networks preferably have an input layer (so-called input layer), at least one hidden layer (so-called hidden layer) and an output layer (so-called output layer).The input layer is preferably used to receive an input.The at least one hidden layer preferably serves for processing the input. In an alternative embodiment, the artificial neural network can also have a plurality of hidden layers. The plurality of hidden layers are preferably cross-linked to each other.The output layer is preferably designed and configured for outputting the processed data.The artificial neural network can be, for example, but not by way of limitation, a perceptron or a multi-layered perceptron or a feed-forward neural network or a recurrent neural network or a modular neural network or a combination of a plurality of the above networks.It can be particularly advantageous if a correlation between at least one area of the vehicle over which the at least one piece of vehicle information is present and at least one area of the vehicle that is vehicle information-free is determined by the artificial intelligence.The term "area via which at least one piece of vehicle information is present" can be understood in the sense of this application as preferably an area of the vehicle and / or of a component of the vehicle on and / or in which at least one sensor is arranged.The at least one sensor preferably acquires the at least one piece of vehicle information of the vehicle and / or of the component of the vehicle. Alternatively, the at least one sensor preferably provides the at least one piece of vehicle information about the vehicle and / or the component of the vehicle.The term "area which is vehicle information-free" can be understood in the sense of this application as particularly preferably to mean an area of the vehicle and / or of a component of the vehicle via which no vehicle information is present.For example, this area does not have a sensor that can be used for capturing and / or providing the at least one piece of vehicle information.The correlation between the two aforementioned ranges is preferably one of the main advantages of the use of AI according to the invention.On the basis of the training of the AI by the at least one piece of vehicle information, the AI can preferably make predictions about at least one piece of vehicle information of the vehicle informationless region.Based on the predicted vehicle information, the AI may then preferably determine one or more operating loads of the vehicle and / or the component of the vehicle.Advantageously, no sensor needs to be arranged in the vehicle information-free area in order to obtain vehicle information. That is to say, vehicle information is preferably also "captured" in regions via which no sensor-supported vehicle information is present. The term "sensor-assisted" can be understood in the sense of this application to mean that the at least one piece of vehicle information is detected and / or transmitted and / or provided by at least one sensor.The advantage of this embodiment can be seen in particular in the fact that the correlation between the two regions preferably creates overall monitoring of the vehicle with respect to at least one piece of vehicle information, without, for example, attaching one or more sensors to each component.The determination of the one or more operating loads is thereby likewise advantageously improved and simplified.It may be advantageous if one or more static operating loads and / or one or more dynamic operating loads are determined. In particular, it is advantageous if load cycles are determined.Preferably, the determination is made by the operating load determination unit. Particularly preferably, the determination is carried out using KL.The one or more static operational loads and / or the one or more dynamic operational loads preferably occur within an operating window of the vehicle.It may be advantageous if the one or more static operating loads and / or the one or more dynamic operating loads are determined within the operating window of the vehicle.The term "operating window" in the sense of this application can preferably be understood to mean a duration of an operation of the vehicle. For example, a vehicle designed as a rail vehicle can involve a trip from one station to the next station in the case of an operating window.Alternatively or additionally, the one or more static operating loads and / or the one or more dynamic operating loads can / can be determined according to an embodiment preferably outside the operating window. With reference to the previous example, this may be the case, for example, when the rail vehicle has finished its travel and / or before it starts its travel.The term "static operating load" can be understood in the sense of this application to mean an operating load which preferably only occurs once and / or which does not change after its occurrence, for example does not increase or decrease in its intensity.Particularly preferably, the static operating load occurs only once in an operating window and / or does not change after its occurrence within the operating window.The term "dynamic operating load" can be understood in the sense of this application as an operating load which preferably occurs several times in an operating window and / or changes after its occurrence, for example increases or decreases in intensity.Preferably, the dynamic operating load occurs multiple times in an operating window and / or changes after its occurrence within the operating windowThe advantage here is to be seen in the fact that not only an amplitude of the operating load, i.e. for example an intensity of a material load, can be determined. Rather, a duration of the load can also preferably be determined and evaluated by the AI.For example, a load with a very high amplitude may have less influence on the vehicle and / or the component of the vehicle than a load with a weaker amplitude, which however occurs very often.The at least one piece of vehicle information is particularly preferably at least one piece of operating and / or technical information about the vehicle.In particular, the at least one piece of vehicle information can be an item of operating and / or technical information about at least one component or a component of the vehicle.The at least one component of the vehicle is preferably, but not restrictive, a chassis and / or a clutch and / or a wheel suspension.It can be advantageous if the at least one component or the component of the vehicle is preferably a part which is exposed to a mechanical load and / or which has a securing and / or protective function.However, the at least one component of the vehicle is not limited to the above examples.According to a general embodiment, the at least one component or component of the vehicle can be any desired component or any desired component of the vehicle.The term "component" can be understood in the sense of this application to mean a plurality of components combined to form a unit. For example, a plurality of axles preferably form a bogie of the vehicle. The axles are here, for example, the components and the bogie is the component preferably formed therefrom.The term "operating information" can be understood in the sense of this application generally and preferably as any information that can occur during the operation of the vehicle.Non-limiting examples of operating information can be: a temperature of the component and / or a rotational speed and / or occurring vibrations and / or a current deflection of a component from a predefined position.The term "technical information" in the sense of this application can preferably be understood to mean any information about the component which can be taken from a data sheet, for example.Non-limiting examples of technical information may be: a material used and / or a material thickness and / or maximum permissible operating parameters.Alternatively or additionally, the at least one piece of vehicle information can preferably also be external influences, such as impacts and / or an impact.It can be particularly advantageous if at least one load collective of the vehicle is determined.The at least one load collective is preferably determined on the basis of the one or more determined operating loads.It can be favorable if at least one load collective of at least one component of the vehicle is determined.The term "load collective" in the sense of this application can preferably be understood to mean the totality of all loads occurring on a component or on a component of the vehicle over a specific period of time.The loads occurring are triggered, for example-but not restrictive-by torques and / or rotational speeds and / or accelerations and / or rotational speeds and / or temperatures.In one embodiment, a plurality of load collectives are preferably determined.By ascertaining the at least one load collective, the quantity of data to be processed is advantageously reduced in comparison to a method which does not use a load collective.This data reduction can preferably also have a positive effect on noise and measurement errors.The at least one ascertained load collective preferably contains information about which load can have the greatest effect on the vehicle or a component of the vehicle.It can be particularly advantageous if, according to a preferred embodiment, a fatigue analysis is carried out on the basis of the at least one load collective determined.The fatigue analysis is preferably carried out by the model of the vehicle.The performance of the fatigue analysis preferably serves to determine and / or predict loads, in particular mechanical loads, of the vehicle.Such ascertainments and / or predictions have proven to be advantageous in the design and / or design and / or optimization of vehicles and in particular of components of vehicles.The object directed with regard to the device is achieved in concrete terms by a device for determining one or more operating loads of a vehicle. In particular, the vehicle is a rail vehicle.The apparatus preferably has an information provision unit and an operating load determination unit.It can be advantageous if the information provision unit and the operating load determination unit are designed and configured in such a way that they carry out the method described above.Preferably, within the scope of this application, a computer program product and a computer-readable storage medium are disclosed and claimed.The computer program product preferably comprises instructions which, when the program is executed by a computer, cause the computer to carry out the method described above.The computer program product can be stored on the device, for example.Preferably, the computer readable storage medium comprises instructions which, when executed by a computer, cause the computer to carry out the method described above.It can be advantageous if the computer-readable storage medium is part of the device.Further preferred features and / or advantages of the invention are the subject matter of the following description and the graphical representation of exemplary embodiments.In the drawings, there are shown: FIG. 1 shows a schematically illustrated block diagram of the method according to the invention for determining one or more operating loads of a vehicle; FIG. 2 shows a schematic illustration of a device according to the invention for determining one or more operating loads of a vehicle, and FIG. 3 shows a schematic illustration of a component of a vehicle, which has an area over which at least one piece of vehicle information is present and an area which is free of vehicle information.Identical or functionally equivalent elements are provided with the same reference numerals in all figures.Referring to Fig. 1, a schematic block diagram of the method of the present invention is shown.As a first step of the method for ascertaining at least one operating load of a vehicle, a model of the vehicle is generated and / or provided 100.The vehicle can be, in particular, a rail vehicle. However, the vehicle is not limited to a rail vehicle. Rather, other vehicle types and categories to which the method can be analogously applied are also conceivable.The model of the vehicle is preferably a model based on a numerical method. In particular, the model is a finite element model.As a next step, the method comprises providing 101 at least one piece of vehicle information of the vehicle.In addition, a database 205 (cf. FIG. 2 ) is created 102.The creation 102 of the database 205 is preferably carried out automatically. The database 205 preferably has sensor data which provide at least one piece of vehicle information of the vehicle. The database 205 is preferably created from simulations of the model, in particular of the finite element model, of the vehicle.Alternatively, the database 205 can also contain vehicle information which has been recorded and / or created, for example, during tests and / or in all-day operation of the vehicle.The database 205 and preferably the sensor data stored on the database 205 are thus preferably used or provided for validating 104 the model of the vehicle.For this purpose, the vehicle information is preferably compared with simulation results of the model. The model is advantageously adapted, if necessary.In the next step of generating and / or providing 100 the model of the vehicle, simulation parameters are preferably input 106 a, 106 b, 106 c.Particularly preferably, an input 106 aof a number of desired simulations and / or an input 106 bof an effect to be simulated on the vehicle and / or an input 106 cof a number of outputs of the model to be called is carried out here.Alternatively, it may prove to be advantageous if an input 106 bof an effect to be simulated on a component of the vehicle takes place.After this automation and validation 104 of the model of the vehicle, training 108 of an artificial intelligence is carried out with the at least one provided vehicle information item.After training 108 the artificial intelligence, the operating load of the vehicle is determined 110 according to the invention. Preferably, a plurality of operating loads are determined.This is preferably carried out on the basis of the at least one provided vehicle information. In this case, the artificial intelligence is particularly preferably used to determine the one or more operating loads.The artificial intelligence is preferably an artificial neural network. Artificial neural networks have preferably proven advantageous with regard to the recognition of complex patterns in large amounts of data.In the next step of the method, a correlation is preferably determined 112 between at least one area 212 (see FIG. 3 ), via which the at least one piece of vehicle information is present, and at least one area 214 (see FIG. 3 ), which is vehicle information-free.The determination 112 is preferably carried out according to the invention by the artificial intelligence.The determination 112 of the correlation preferably ensures overall monitoring of all regions of the vehicle, even if there is no vehicle information present over some regions of the vehicle. This "gap" is preferably closed by the artificial intelligence and its prediction. Thus, the one or more operating loads may be determined more accurately.Preferably, furthermore, one or more static operating loads are ascertained 114.Alternatively or additionally, one or more dynamic operating loads, in particular load cycles, are preferably ascertained 114.The determination 114 of static and / or dynamic operating loads preferably serves the purpose that different types of influence on the vehicle can be determined in the form of operating loads.The determined operating load or the determined operating loads are output by the artificial intelligence.The method preferably comprises the step of creating 116 at least one load collective of the vehicle.It can be particularly advantageous if the at least one load collective is created on the basis of the one or more determined operating loads.In a step of the method which preferably follows it, a fatigue analysis is preferably carried out 118.The fatigue analysis is preferably carried out 118 on the basis of the at least one load collective determined.It may be advantageous if the fatigue analysis is carried out 118 on the basis of a plurality of load collectives determined.The fatigue analysis is furthermore preferably carried out 118 by the model of the vehicle.The fatigue analysis preferably serves for ascertaining and / or predicting loads, in particular mechanical loads, of the vehicle.In particular, according to an alternative embodiment, the fatigue analysis can preferably also serve for ascertaining and / or predicting loads, in particular mechanical loads, of a component of the vehicle.It may be particularly advantageous if the fatigue analysis carried out, in particular the results of the fatigue analysis, is used for optimizing 120 the vehicle, in particular a component of the vehicle.On the basis of the determined operating loads and / or the generated load collective and / or the fatigue analysis carried out, the vehicle, in particular components of the vehicle, can preferably be adapted in such a way that over-dimensioning is prevented.FIG. 2 shows a schematic illustration of an embodiment of a device 200 for determining 110 one or more operating loads of a vehicle.The vehicle is, as in FIG. 1, likewise-but not restrictive-a rail vehicle.The vehicle is represented in FIG. 2 merely by a component 202 of the vehicle. The component 202 is preferably a part of the vehicle that has one or more components.In the exemplary embodiment according to FIG. 2, the component 202 has only one component. For this reason, the two terms "component" and "component" are used interchangeably below with reference to FIG. 2, and without impairing the definition already mentioned in the general part of the description.The device 202 according to FIG. 2 is preferably designed and configured in such a way as to carry out the method described above for determining one or more operating loads of a vehicle.For this purpose, the device 202 preferably has an information provision unit 204. The information provision unit 204 preferably serves for providing 101 at least one piece of vehicle information of the vehicle.Furthermore, in the exemplary embodiment according to FIG. 2, the information provision unit 204 has a database 205.Particularly preferably, the information provision unit 204 provides at least one piece of vehicle information to the component 202 of the vehicle.For this purpose, the device 202 preferably has sensors 206. The sensors 206 are preferably arranged on the component 202 and according to FIG. 2. Alternatively, the sensors 206 may also be arranged in the component 202.Furthermore, alternatively or additionally, it can be advantageous if at least one piece of vehicle information of the vehicle is stored on the database 205, which is used for the provision by the information provision unit 204.Preferably, the sensors 206 detect loads, in particular mechanical loads, of the component 202. For example, the sensors 206 of FIG. 2 sense strain of the component 202. Advantageously, the sensors 206 according to FIG. 2 are therefore preferably designed as strain gauges.According to an alternative embodiment, the sensors 206 can also be designed as temperature sensors or as acceleration sensors in order to detect different loads of the component 202.Likewise, according to another embodiment, one sensor 206 is, for example, a temperature sensor for detecting a temperature of the component 202 and another sensor 206 is a strain gauge for detecting a strain of the component 202. Thus, a plurality of vehicle information items may be advantageously acquired via a component 202 and supplied to the information provision unit 204.In order to enable a data exchange between the sensors 206 and the device 200, and in particular with the information provision unit 204, the sensors 206 are each preferably connected to the device 200 by means of a line 208.According to an alternative embodiment, which is not shown, the data exchange between the device 200 and the sensors 206 can preferably also take place by means of a wireless data connection.Furthermore, the device 200 according to FIG. 2 has an operating load determination unit 210. The operating load determination unit 210 is preferably used to determine 110 the one or more operating loads of the vehicle.The determined operating loads are preferably based on the at least one provided vehicle information item of the information provision unit 204.For this purpose, the two units 204, 210 can preferably communicate with one another.Moreover, the device 200 is designed and configured for carrying out the method already mentioned above and in particular the steps of the method described above, so that reference is made to the above explanations in this respect.Referring to FIG. 3, a component 202 of the vehicle is shown.The component 202 according to FIG. 3 is configured in an angular manner. It has an area 212 over which at least one piece of vehicle information is present. The area 212 over which the at least one piece of vehicle information is present is graphically highlighted in FIG. 3 by a dashed rectangle.The at least one piece of vehicle information about the area 212 is obtained by the information provision unit 204 (not illustrated in FIG. 3 ) through a sensor 206.In the exemplary embodiment according to FIG. 3, the sensor 206 is arranged on the component 202 and preferably within the region 212.Furthermore, component 202 according to FIG. 3 has an area 214 that is vehicle informationless.The area 214 that is vehicle information-free is also graphically illustrated in FIG. 3 by a dashed rectangle.As can be seen from FIG. 3, the area 214 that is vehicle information-free does not have a sensor 206. This can be due, for example, to the fact that no sensors 206 can and / or may be attached in this region 214 for structural and / or safety reasons.Therefore, the information providing unit 204 preferably does not have any information detected by a sensor 206 about this area 214.FIG. 3 illustrates the advantage of the method according to the invention and of the device 200 according to the invention.By using artificial intelligence, it is advantageously possible to predict at least one piece of vehicle information even via the area 214 that is vehicle information-free.Based on the predicted vehicle information, one or more operating loads can then preferably also be determined for this region 214.For this purpose, the artificial intelligence which is used by or in the operating load determination unit 210 preferably uses the vehicle information of the area 212 determined by the sensor 206 and "transmits" it-optionally adapted-to the area 214.Due to the configuration of the artificial intelligence, it is preferably made possible to establish a correlation between the two areas 212, 214 and thus likewise predict one or more operating loads in the area 214 which is vehicle information-free.List of reference characters100 Generating / providing a model of the vehicle 101 Providing at least one piece of vehicle information 102 Creating a database 104 Validating the model of the vehicle 106 a- c Eingabe simulation parameters 108 Training the artificial intelligence 110 Determining the one or more operating loads 112 Determining a correlation between regions of the vehicle 114 Determining static and / or dynamic operating loads 116 Creating at least one load collective 118 Carrying out a fatigue analysis 120 Optimizing the vehicle, in particular a component of the vehicle 200 Device 202 Component of the vehicle 204 Information provision unit 205 Database 206 Sensors 208 Line 210 Operating load determination unit 212 Region over which vehicle information is present 214 Region which is free of vehicle information

Claims

Computer-implemented method for determining one or more operating loads of a vehicle, in particular of a rail vehicle, the method comprising the steps of: - generating and / or providing (100) a model of the vehicle; - providing (101) at least one piece of vehicle information of the vehicle by an information provision unit (204); - determining (110) the one or more operating loads of the vehicle by an operating load determination unit (210) on the basis of the at least one piece of vehicle information provided, wherein an artificial intelligence is used for determining the one or more operating loads.Method according to Claim 1, characterized in that a numerical method is used for the generation and / or provision (100) of the model of the vehicle, in particular in that the model of the vehicle is a finite element model.Method according to Claim 1 or 2, characterized in that the information provision unit (204) has a database (205) or is designed as a database (205), wherein the at least one piece of vehicle information is provided by the database (205); wherein the following step can preferably be provided: - creating (102), in particular automatically creating, the database (205) when generating and / or providing (100) the model of the vehicle.Method according to one of the preceding claims, comprising the step of: - generating and / or providing (100) the model of the vehicle by one or more of the steps: - inputting (106a) a number of desired simulations; and / or - inputting (106b) an effect to be simulated on the vehicle; and / or - inputting (106c) a number of outputs of the model to be called.Method according to one of the preceding claims, characterized in that - the information provision unit (204) has at least one sensor (206), in particular a plurality of sensors (206), which is arranged on and / or in the vehicle, wherein the at least one piece of vehicle information is provided by the at least one sensor (206); and / or - the information provision unit (204) uses one or more sensors (206) of the vehicle to provide the at least one piece of vehicle information.Method according to claim 5, characterised in that the at least one sensor (206) is one of the following sensors: - force sensor; and / or - acceleration sensor; and / or - shock absorber; and / or - strain sensor; and / or - pressure sensor.Method according to one of the preceding claims, in particular according to Claim 5 or 6, further comprising the step of: - validating (104) the model of the vehicle by means of the at least one piece of vehicle information detected by the at least one sensor (206).Method according to one of the preceding claims, further comprising the step of: - training (108) the artificial intelligence with the at least one provided vehicle information.Method according to one of the preceding claims, characterized in that the artificial intelligence is designed as an artificial neural network.Method according to one of the preceding claims, further comprising the step of: - determining (112), by the artificial intelligence, a correlation between at least one area (212) of the vehicle over which the at least one piece of vehicle information is present and at least one area (214) of the vehicle which is vehicle informationless.Method according to one of the preceding claims, further comprising the step of: - determining (114), by the operating load determination unit (210), one or more static operating loads and / or one or more dynamic operating loads, in particular load cycles, using the artificial intelligence.Method according to one of the preceding claims, characterized in that the at least one piece of vehicle information is at least one piece of operating and / or technical information about the vehicle, in particular about at least one component (202) of the vehicle.Method according to one of the preceding claims, further comprising the step of: - creating at least one load collective of the vehicle, in particular at least one component (202) of the vehicle, on the basis of the one or more determined operating loads.Method according to one of the preceding claims, in particular according to claim 13, further comprising the step: - carrying out (118) a fatigue analysis on the basis of the at least one ascertained load collective by the model of the vehicle for ascertaining and / or predicting loads, in particular mechanical loads, of the vehicle.Device (200) for determining one or more operating loads of a vehicle, in particular of a rail vehicle, the device (200) having: - an information provision unit (204) and - an operating load determination unit (210); wherein the information provision unit (204) and the operating load determination unit (210) are designed and configured in such a way that they carry out the method according to one of Claims 1 to 14.

Images